2.3 The Norwegian health care context
The Norwegian Coordination reform (7) was addressing the fragmentation in care and the barriers to collaboration between hospitals, GPs and other primary care services related to ownership, management, funding, legislation, information systems, and professional aims.
Most hospitals, which are responsible for specialist care, are owned and funded by the Government and managed by four regional health authorities as public enterprises with a steering group and a chief executive officer (100). A few small, private owned hospitals operate on contract with the regional health authorities. The local municipal authorities (municipalities) are responsible for providing primary health and social care.
This includes regular GPs who usually run private enterprises on contract with the municipality. In contrast to hospitals with line management from the Ministry of Health, a local democratically elected council is responsible for priorities and funding of all services in the municipalities. The municipalities have a large degree of freedom on how they want to prioritise and organise their services. This implies that planning and priorities of health care services at primary and secondary level are not always attuned.
Separate laws to specialist care and primary care reinforces this tendency.
27
Separate electronic health record systems (EHRs) are also an obstacle to coordinated care. Shared health records between different organisational units (e.g. GP practices, hospitals, home care services) have been illegal in Norway due to strict rules for protection of privacy. At the time of the introduction of the Coordination Reform, electronic exchange of information was limited to referral and discharge letters between the GPs and the hospitals and some test results from hospitals to the GPs. This made communication between collaborating partners ineffective as they were working at different locations (in most cases) and had to communicate by post and by fax or phone in urgent cases. A new legislation about EHR and electronic communication (101) came into force in January 2015 implying that sharing EHR systems across organisations and exchange of patient information will be easier ahead.
Finally, different professional aims and perspectives represent barriers for integrated care. In hospitals, there is increasing specialisation of services and focus on efficient trajectories and high quality, evidence-based treatment of the diseases causing the admittance (9). The main aim in primary health and social care services is to support people to manage their life as independent as possible and maintain quality of life taking into account functional ability, social situation and care for all the diseases the patients might have; thus caring for the person rather than caring for the disease. Such different perspectives and aims represent an obstacle to information flow as one party might not know or take into account the information needed by collaborating partners (102, 103).
In this context, the framework of an integrated cross-boundary care pathway was outlined by healthcare managers in the City of Trondheim in cooperation with St. Olavs University Hospital and researchers from Department of Public Health and General Practice at NTNU. It was primarily a quality improvement project addressing elderly people in need of home care services after hospital discharge. The care pathway, which was named Patient Trajectory for Home-dwelling elders (PaTH), was to include discharge arrangements and post-discharge follow- up in order to improve coordination and quality of care and thus support elderly patients to stay in their homes as long as
28
possible with high quality of life. It became part of the project “Samhandlingskjeden kronisk syke” (Integrated Care Services for the Chronically Ill Patient) (104). The management level in six municipalities and three hospitals in the Central Region of Norway agreed to take part in the project. Furthermore, it was decided to do a scientific study of the development and implementation process and to analyse effectiveness on the target population (home care recipients).
29
3 Objectives, aims and hypotheses
The objective of this thesis was to provide knowledge about the development of an integrated care pathway for elderly patients involving specialist and primary care services, and furthermore, knowledge about its implementation and effectiveness. This was investigated by conducting studies with the following aims:
Paper 1: To explore the process of developing the integrated care pathway that was going to be implemented in the project.
Paper 2: To investigate the process of implementing the integrated care pathway into everyday practice by comparing the joint experiences of health care professionals and managers in home care services between the municipalities where it had been introduced.
Paper 3: To establish the effectiveness of PaTH on patient level – compared to usual care – for elderly patients in need of home care services after discharge from a general hospital, regarding primarily the patients’ functional level and readmissions, secondarily use of health care service, mortality and quality of life.
The following hypotheses were defined for Paper 3:
An integrated care pathway targeting elderly home care recipients will result in:
H1: Decline in the patients’ functional level (ADL scores) will be reduced H2: Health care utilisation will be reduced
x Number of readmissions to general hospital will be reduced
x Number of hospital admissions and days of inward hospital care will decrease x Need for long-term institutional care in the municipalities will be postponed x Number of admissions to and days in short-term and long-term community
institutional care will decrease H3: Number of GP consultations will increase H4: Mortality will be reduced
H5: Patients will get more days at home
30
H6: Patients will get higher level of health related quality of life
31
4 Material and methods
Evaluation of complex interventions should include randomised controlled trials to be able to conclude on effect (105). However, evaluation is demanding, and process evaluation is considered an essential part of designing and testing complex interventions (106, 107). The UK Medical Research Council (108) recommends use of both quantitative and qualitative research methods; quantitative research methods alone are considered insufficient (108) as they only answer questions that can be measured and counted (109). Evaluation of a complex intervention like care pathways, also need to explore questions like: what are the active ingredients in the care pathway, how are they excreting their effects, how feasible is the intervention in daily work, was the intervention implemented and delivered as intended, and what were the reasons for the findings of the trial (110, 111). These questions can best be answered by qualitative research methods which aim to explore, interpret, or obtain an in-debt understanding of the phenomenon or the attitudes, expectations, motivation, or experiences of particular individuals or groups (112, 113). Qualitative and quantitative research strategies can thus be seen as complementary (114).
This thesis therefore comprises three integrated studies, presented in three papers: a cluster randomised controlled trial with nested qualitative studies of the development and
implementation of the care pathway.
The study was approved by the Regional Committee for Medical and Health Research (REK 4.2009.670) and the Ombudsman for Research and Social Science Data Service (NSD 215289). The trial was registered in Clinical Trials.gov NCT01107119. All informants in the process evaluations and patients in the cRCT were informed about the study and its purpose, that their anonymity would be ensured and that they could withdraw from the study at any time. They all signed an informed consent before participation. We did not consider the intervention or study to represent potential harm to patients or informants.
32 4.1 Qualitative studies
Study 1. Development of a patient-centred care pathway across health care providers: a qualitative study
Study 2. Implementing a care pathway for elderly patients, a comparative process evaluation in primary care
4.1.1 Study design and material
A qualitative design was used in the development study and the implementation study in order to explore the experiences, expectations, motivations, and attitudes of the health care
professionals involved in developing and implementing the integrated care pathway.
Interviews were the main data source and minutes, notes, and observations were mainly used to supplement the interviews, assisting in analysing the meaning of the informants’
experiences.
In the development study data were collected through
x observations of discussion and work in regional and local working groups x minutes and notes from working groups
x individual and focus group interviews performed after the development process
In the implementation study data were collected through
x minutes from telephone conferences during the early implementation process
x individual and focus group interviews during early implementation process and two and three years after the care pathway was introduced in the municipalities
x review of the electronic patient records two years after the introduction of PaTH in order to assess how the checklists were integrated in the electronic health records
The interviews were semi-structured; the interview guides had open-ended questions allowing the participants large degree of freedom about what they would talk about and emphasise (Table 1).
33 Table 1. Topics included in the interview guides
Development study
Understanding of the care pathway concept Responsibilities and cooperation in a care pathway Expectations and attitudes during the development process Challenges in the development process
Other important issues appearing during the development work Appraisal of the final version of the care pathway
Implementation study
How the informants had been involved with the care pathway Their initial expectations to the care pathway
How it had been introduced at their workplace The efforts invested to take it into use
Challenges and promoting factors during implementation Assessments of benefit
If and why it was dismissed or integrated and sustained in daily use
4.1.2 Informants
Twenty-three informants (four focus groups) were recruited to the development study (Paper 1), and 60 informants (12 focus groups) were recruited to the implementation study (Paper 2) (Table 2). In addition, there were some individual interviews of informants who were not able to participate in the focus group interviews or informants who belonged to the management level. In Paper 1, the focus groups had four to seven informants and the interviews were held (except for the GP group) at each of the three hospitals. The groups (except for the GP group) were mixed with informants from the hospitals and adjacent municipalities (Table 2 in Paper 1). In Paper 2, the focus groups had three to eight informants. The interviews in 2012-2013 were performed at the local workplaces of the home care professionals.
34
Table 2. Participants in the interviews 2010 and 2011-2013 Focus
2010 Mixed primary and specialist care and patient associations
2012 Management level primary care Frontline nurses and nursing assistants primary care 2013 Management level primary care
Frontline nurses and nursing
The researcher recruited the informants to the development study using two selection criteria; the informants should have participated actively throughout the development project, and all hospitals and municipalities should be represented by at least two participants. The informants represented two patient associations, five of the six municipalities, the three hospitals, and the Central Norway Regional Health Authority.
One small municipality was not represented in the interviews due to problems with capacity, and they temporarily pulled out of the project. The informants constituted about half of those that had been active in the regional and local workgroups. One nurse was not able to participate in the focus group and was interviewed individually to ensure at least two representatives from each hospital. Few GPs took part in the development process, and none of them were able to take part in the mixed focus group interviews. Collaboration between home care services and GPs was, however, an important topic both in the development process and in the interviews. Therefore, an additional focus group of four GPs was recruited to elaborate collaboration from their perspective as physicians. Of these four, only one had taken part in the actual development process.
35 The implementation study
The care pathway (PaTH) was to be implemented in the three hospitals and six
municipalities (including 48 GP practices) that had taken part in the development work.
Due to available time and resources, the study of the implementation process was limited to the home care services. The researcher recruited every home care manager and head nurse in the home care units where PaTH had been introduced. The head nurses / home care managers recruited in turn frontline nurses and nursing assistants ensuring both professional groups from all municipalities. To be selected, the informants should have been employed in the home care services from the first introduction of PaTH and have experience in using the checklists.
In both the qualitative studies, the interviews lasted for one to two hours, were audio recorded and transcribed verbatim. The researcher performed all the interviews and transcribed half of the interviews. The rest was transcribed by research assistants, but the researcher checked all transcripts against the audio-file.
4.1.3 Analytical method and theoretical framework
The interviews in both studies were analysed using systematic text condensation. In the implementation study, the Normalisation Process Theory was applied to systematise the comparison between the municipalities in order to facilitate understanding of the factors of importance as to why PaTH was integrated and sustained in the home care services in some municipalities and not in others.
Systematic text condensation
Systematic text condensation (STC) was developed by Malterud (115) and inspired by Giorgi’s psychological phenomenological approach (116). Phenomenological analyses seek to catch the meaning and essences of an event through examination of individual experiences, which was the purpose of the qualitative studies. STC represent a pragmatic approach, applicable for use with different theoretical frameworks.
36
The data from the interviews in the qualitative studies were organised and analysed according to the four steps of STC (115). In the first step, the research group read the whole material to get an overall impression, and the main researcher identified preliminary themes. In the second step, the researcher reviewed the text systematically line for line and identified units of meaning, which is text fragments containing information about the research question. The meaning units were classified into code groups, mainly related to the preliminary themes. The whole research group
reconsidered and refined the code groups. In the third step, the meaning units in the code groups were further classified into subgroups. Furthermore, a condensate was made, representing the content of all meaning units in one subgroup, and illustrative quotations were identified for each subgroup. In the fourth step, the researcher wrote an analytic text based on the condensates and quotations, presenting the most salient content and meaning of the phenomenon under study.
Normalisation process theory
In the implementation study (Paper 2), we searched for a theory explaining what is necessary for implementing and integrating the care pathway into daily working practices. We also wanted a theory that could make a framework for comparing the participating municipalities. Among several implementation theories, we found the Normalisation Process Theory (NPT) to be most applicable to our needs. NPT was developed and presented by Carl May, Tracy Finch and colleagues in 2009 as a framework for understanding processes by which complex interventions become, or do not become, part of normal practice (117). Initially NPT was developed from studies of health technology implementation (118), but has during the last years been tested, refined and applied in studies conducted across diverse settings in hospitals as well as in primary care (118, 119). The theory is useful for evaluating implementation of complex intervention as well as to design and judge the implementation potential of new interventions (120). The focus of NPT is not only on early implementation, but beyond this to a point where the intervention is integrated and sustained as a natural part of daily work (120). This makes the theory useful alongside randomised controlled trials to assess whether an intervention is feasible and sustainable beyond the project setting.
37
Four core constructs define essential conditions and processes for new working practices to become a natural part of daily work: coherence (making tasks meaningful and understandable), cognitive participation (building commitment and engagement), collective action (efforts and resources invested to make the intervention work), and reflexive monitoring (assessment of benefit). These four constructs of the NPT framework were used to map the subcodes in the last step of the STC process. The relationship between the constructs is not linear, they influence each other mutually, and implementation work is necessary within all four constructs for a complex intervention to be integrated and sustained in daily work (117).
4.2 Cluster randomised controlled trial
Study 3: Generic care pathway for elderly patients in need of home care services after discharge from hospital: A cluster randomised controlled trial
4.2.1 Study design
A cluster randomised controlled trial (cRCT) design was chosen to study the
effectiveness of the integrated care pathway. The home care service units were the unit of randomisation and were randomised to either use the intervention or their usual procedures (control group). Outcomes were assessed at the level of the individual home care recipients.
Senior medical officers in an independent organisation (Ministry of Health) were drawing lots, identifying one home care cluster in every municipality to start as intervention clusters. Initially a step-wedge design (121) was planned in each
municipality. The plan was that the control clusters should gradually implement the care pathway in a predetermined manner. To avoid contamination of the control patients, inclusion of patients was planned to be stopped two months before introduction of the care pathway. However, implementing PaTH was experienced to be far more
demanding than expected, and except for one cluster in Trondheim, all the initial control units remained controls throughout the study period of the cRCT.
38
In order to test feasibility and acceptability of the intervention, a test period of three months, involving one home care unit, was preceding the main trial. Only some minor details were changed before it was introduced to all intervention clusters.
4.2.2 Clusters and study population
Twelve clusters were formed, each consisting of one to three home care units. The total number of home care recipients in the different clusters varied considerable, ranging from 80 to 750. Eligible patients were at least 70 years of age and in need of home care services after discharge from hospital. They were to be discharge directly to their homes or via a transitional stay in a short-term community institution anticipated to be less than four weeks. Exclusion criteria were participation in another intervention study or cognitive impairment to such an extent that the patient was not able to sign an informed consent.
From January to April 2010, after the test period, patients were included in the trial successively at all intervention sites after a one-day introduction course to all
employees. The patients were recruited in either a discharge meeting at the hospital by municipal case managers (in Trondheim) or immediately after returning home by a home care nurse (all municipalities).
4.2.3 Outcomes and data collection
Functional ability was chosen as one of our two primary outcomes as disability has major implications in older patients regarding quality of life, need for support and need for long-term care, either at home or in long-term institutions in the municipalities.
Functional ability is generally assessed through self-report or proxy report on
difficulties or need for help related to basic self- care tasks, mobility or more complex tasks for living independently in the community (115). Most commonly used are measures of daily living (ADL) and instrumental activities of daily living (IADL) (115).
ADL is related to basic self-care tasks and include bathing, dressing, transferring form a
ADL is related to basic self-care tasks and include bathing, dressing, transferring form a